Production of biodiesel from acid oil using sulfuric acid as catalyst: kinetics study

Transcription

1 International Journal of Low-Carbon Technologies Advance Access published October 15, 2010 *Corresponding author: ntnu.no Production of biodiesel from acid oil using sulfuric acid as catalyst: kinetics study... J.M. Marchetti 1 *, M.N. Pedernera 2 and N.S. Schbib 2 1 Chemical Engineering Department, Faculty of Natural Science and Technology, Sem Sælands v.4 NO-7491, Trondheim, Norway; 2 Planta Piloto de Ingeniería Química (Universidad Nacional del Sur CONICET), Camino La Carrindanga km. 7, 8000 Bahía Blanca, Argentina... Abstract Biodiesel production is becoming more relevant due to its environmental advantages together with the global petroleum situation regarding decreases of the reserves, instability of the prices and so on. The kinetics of the esterification and transesterification reaction of an acid oil (10% of initial free fatty acid) was studied using ethanol and a homogenous acid catalyst. The controlling step method was employed to obtain the kinetic equations, and nonlinear multiparametric routine was implemented to estimate the kinetic parameters. The selected reaction rate represented the experimental data for several operating conditions satisfactorily. Keywords: biodiesel; esterification; kinetic; sulfuric acid; acid oil Received 9 July 2010; revised 8 September 2010; accepted 13 September INTRODUCTION Fuel consumption has considerably increased over the last few years; therefore, great efforts have been made to find other renewable and sustainable alternatives for fossil fuel. Biodiesel is a valuable option because of its environmental benefits such as lower toxicity for humans, i.e. lower CO and sulfur emissions and a minimum amount of particle matter [1 4]. The most common way to produce biodiesel is by transesterification of vegetable oils with an alcohol. The reaction can be catalyzed by alkalis, acids or enzymes or by carried out under supercritical conditions [5 21]. CH 2 OOC R CH OOC R CH 2 OOC R þ 3R 0 OH catalyst! The main disadvantage in biodiesel production is the high cost of the raw materials used, normally refined oils. Therefore, different feedstocks (waste, acid and frying oils) are being evaluated as possible substitutes. These alternative raw materials are less expensive, although being likely to appear more impure. In acid oils, the amount of free fatty acid (FFA) could vary from 3 to 40%. When the amount of FFAs in the feedstock exceeds 0.5%, the use of the conventional alkali catalyst is not recommended due to the fact that the saponification reaction might take place [2 4]. Nevertheless, the FFA could be transformed into ester by means of the direct esterification reaction. The main reactions involved when an acid oil is used are summarized as follows: R 1 OOC R 0 R 2 OOC R 0 R 3 OOC R 0 þ CH 2 OH CH OH CH 2 OH Triglycerides Alcohol Esters Glycerin ð1þ Downloaded from at Pennsylvania State University on March 6, 2016 R COOH þ Free fatty acid R0 OH Alcohol catalyst! H 2 O þ R CO OH 2 C R 0 Water Easter ð2þ where R and R 0 denotes any hydrocarbon chain. International Journal of Low-Carbon Technologies 2010, 0, 1 6 # The Author Published by Oxford University Press. All rights reserved. For Permissions, please ournals.permissions@oxfordournals.org doi: /ilct/ctq040 Advance Access Publication 00 Month of 6

2 J. M. Marchetti et al. The kinetics of the esterification reaction has extensively been studied [22 30]. However, in those works, the FFA was considered a pure component. Berrios et al. [31] and Tesser et al. [32] have considered acid oil with several concentration of FFA. Nevertheless, the amount of triglycerides has not been included in the reaction mechanism proposed. It could be considered constant if its concentration does not change during the reaction. However, this is unlikely to happen because the conditions for the transesterification to take place are appropriate. In the present work, the kinetics of the direct esterification of the FFA has been studied using acid oils in two different contexts. In the first case, triglycerides have been considered inert. Consequently, only the esterification reaction (2) has been considered. From laboratory results, a second alternative was proposed where the transesterification reaction of triglycerides (1) occurs together with the direct esterification reaction (2). For both cases, the kinetic models were developed and their kinetic parameters have been fitted using experimental data obtained at different operating conditions such as reaction temperature, amount of catalyst, alcohol/oil molar ratio and different initial percentages of FFA in the mixture. These kinetic models were obtained applying the rate-controlling step method to the different reaction mechanisms proposed. Both alternatives were found to be suitable to reproduce the experimental data for different temperatures and amounts of catalyst. However, when both reactions are taken into account, the kinetic model fits the information more accurately. 2 MATERIALS AND EXPERIMENTAL METHOD The experimental runs were carried out in a lab scale reactor of 500 ml equipped with a mechanical agitation manipulated by speed control at 200 rpm. The reaction temperature was controlled by thermostatic bath that allowed setting the temperature at the desirable value with an error of +0.18C. To carry these experiments, refined sunflower oil from a commercial brand were used. Ethanol anhydrous and sulfuric acid, brand Anedra, were utilized as reagent and catalyst, respectively. Ethanol was chosen instead of methanol due to its lower toxicity and the fact that this alcohol could be produced from renewable sources. To obtain an oil acid of a known acidity, an amount measured of oleic acid was added to the vegetable oil. The mixture was prepared and fed to the reactor to be heated up. Once the reaction temperature was reached, the catalyst mixed with the alcohol was fed and the reaction started. In each run, the liquid samples were withdrawn at different specific times by means of a valve placed at the bottom of the reactor. Once it was removed from their reaction mixture, the sample was washed with water in order to stop the reaction as well as to facilitate the separation. To improve the separation of phases, a centrifuge was used for 20 min. From the upper phase (oil phase), a small sample was taken for GC analysis and a slightly bigger sample was used for titration. Titration procedure was done by diluting the oil sample into ethyl ether and ethanol (50% vol), adding phenolphthalein as an indicator. The acidity was then measured by using a KOH solution of concentration and normality known. This was done accordingly to AOCS Standard Methods Ca 5a-40 [33]. The amount of KOH consumed was recorded, and the acidity of the sample could be calculated by means of the following relation: a ¼ V 100 PM C Mg From to the acidity of the system, the amount of oleic acid (C ffa ) is therefore established at each time. Consequently, the conversion of FFA (x ffa ) was calculated using this expression: x ffa ¼ at¼0 ffa at¼t ffa affa t¼0 where affa t¼0 is the initial acidity of the fatty acid mixture and affa t¼t is the acidity of the same at t time. In this case, that sulfuric acid was used as catalyst, it was taken care that the acid was not present in the oil phase. The experimental data selected by this study showed that sulfuric acid does not appear in the samples within of 1.8% in the acid index. The concentration of triglycerides, diglycerides, monoglycerides, biodiesel and glycerin were analyzed by capillary gas chromatography (GC). Each sample of the oil phase was diluted in pyridine (50 mg/5 ml), 100 ml from of the solution were mixed with 200 ml N,O-bis(trimethylsilyl)-trifluoroacetamide and 100 ml trimethylchlorosilane. This new mixture was heated up for 20 min at 708C and 1 ml was then inected to a GC Varian The GC was fitted with 15 m 0.32 mm 0.1 mm column DB-5HTJ&W Scientific. The titration and the GC method are complimentary methods due to the fact that different compounds could be analyzed with this technique. However, the FFA was checked with both techniques. This procedure allowed us to verify the results of the GC analysis. 3 KINETIC MODELS The kinetic expressions were obtained using the controlling step reaction method applied to several possible reaction mechanisms. The FFA conversion as a function of time was calculated using the batch reaction equation: dx ffa ¼ r ffa C 0 ffa The kinetic parameters from the proposed mechanism were ð3þ ð4þ ð5þ Downloaded from at Pennsylvania State University on March 6, of 6 International Journal of Low-Carbon Technologies 2010, 0, 1 6

3 Production of biodiesel from acid Table 1. Kinetic parameters. Parameter Value Units k l/(kg s) E kj/mol K DH kj/mol K l 2 /mol 2 K l/mol K l/mol adusted with experimental data presented elsewhere [34,35]. To carry out this procedure, a non-linear regression routine based on the Marquard [36] algorithm was used. The selection of the model was done by comparing statistical criteria such as standard deviation and variance among others. 3.1 Case I: esterification reaction In this case, only the proposed esterification reaction took place whereas the transesterification reaction did not. Hence, the triglycerides can be considered inert in the reaction. To obtain the kinetic expression, the mechanism proposed by Straitweiser et al. [37] was chosen. From this mechanism and applying the controlling step method, several possible expressions for the kinetic reaction rate were obtained. Upon applying statistical criteria, the following reaction rate (r, mol/ l/s) turned out to be the most accurate one. m 0 cat r ffa ¼ 1 þ K 1 C faee C w þ ðk 2 C faee C w =C ffa ÞþK 3 C faee k C ffa C ð6þ faeec w C alc K where m 0 cat is initial concentration of the catalyst (kg/l), faee fatty acid ethyl oleate, C concentration of a compound (mol/l), k kinetic constants in l/(mol h), k ¼ k 1 exp ð E=RTÞ, where E is activation energy (kj/mol), R is universal gas constant and T is temperature (K), andk ¼ K 1 exp ð DH=RTÞ, where K i is equilibrium constants and DH is heat of reaction (kj/mol). The kinetic parameters for the selected model are reported in Table 1. The obtained activation energy (E ¼ kj/mol) is of the same magnitude as one obtained by Sendzikiene et al. [30] ( kj/mol). The experimental and calculated conversion of FFA for three different reaction temperatures (35, 45 and 558C) are shown in Figure 1. As it can be seen, the model predicts the experimental data satisfactorily. The effect of the amounts of catalyst is displayed in Figure 2. Figure 1. Conversions of FFA for different temperatures. Dashed line area of the curve denotes the simulated values and filled symbols experimental data. Case I: initial FFA ¼ 10.68%; n ¼ 6.1:1; S ¼ 2.1%; rpm ¼ 200. Figure 2. Conversions of FFA for different values of S. Dashed line area of the curve denotes the simulated values and filled symbols denote experimental data. Case I: initial FFA ¼ 10.68%; n ¼ 6.1:1; rpm ¼ 200; T ¼ 458C. 3.2 Case II: esterification and transesterification reactions Considering experimental laboratory results, the transesterification reaction could take place simultaneously with the esterification reaction. For this reason, a global reaction mechanism should be proposed to obtain the kinetic expression for both the esterification and the transesterification reactions. The proposed mechanism is as follows: (1) TG þ ALC $ DG þ FAEE (2) DG þ ALC $ MG þ FAEE (3) MG þ ALC $ G þ FAEE (4) FFA þ ALC $ H 2 O þ FAEE Each step in the above mechanism is an elementary reaction. Downloaded from at Pennsylvania State University on March 6, 2016 International Journal of Low-Carbon Technologies 2010, 0, of 6

4 J. M. Marchetti et al. Figure 3. Conversions of FFA for different temperatures. Dashed line area of the curve denotes the simulated values and filled symbols denote experimental data. Case II: initial FFA ¼ 10.68%; n ¼ 6.1:1; S ¼ 2.1%; rpm ¼ 200. The variations in the concentration of each compound is defined as follows dc faee dc tg ¼ r 1 dc dg ¼ r 1 r 2 dc mg dc g ¼ r 2 r 3 ¼ r 3 ¼ r 1 þ r 2 þ r 3 þ r 4 dc alc dc ffa ¼ dc faee ¼ r 4 ð7þ ð8þ ð9þ ð10þ ð11þ ð12þ ð13þ For the esterification reaction, the same model from Case I was proposed. For the transesterification reaction, a new kinetic expression was obtained, assuming that all steps were elementary. The following resulting expressions were attained: r 1 ¼ k 1 C tg C alc k 2 C dg C est r 2 ¼ k 3 C dg C alc k 4 C mg C est r 3 ¼ k 5 C mg C alc k 6 C g C est m 0 cat ð14þ ð15þ ð16þ r 4 ¼ 1 þ K 1 C faee C w þðk 2 C faee C w =C agl ÞþK 3 C est k 7 C ffa k ð17þ 8C faee C w C alc where k i ¼ k i1 exp ð E=RTÞ ; i ¼ 1; 8: The experimental conversion of FFA and those from the kinetic model can be observed in Figure 3 for the same Figure 4. Concentrations of TG for different temperatures. Dashed line area of the curve denotes the simulated values and filled symbols denote experimental data. Case II: initial FFA ¼ 10.68%; n ¼ 6.1:1; S ¼ 2.1%; rpm ¼ 200. Figure 5. Experimental and calculated FAEE concentrations. Dashed line area of the curve denotes the simulated values and filled symbols denote experimental data. Case II: initial FFA ¼ 10.68%; n ¼ 6.1:1; S ¼ 2.1%; rpm ¼ 200. temperatures of Case I. Once again, a very good representation of the experimental data is achieved. The evolutions of the concentration of TG and biodiesel as a function of time are shown in Figures 4 and 5. As shown, when the temperature increases, the TG concentrations decreases and the biodiesel production increases. The accuracy of the kinetic model for three different amounts of catalyst is presented in Figure 6, considering the variations of the FFA conversions. The kinetic parameters obtained from the fitting of the experimental data with the Case II are stated in Table 2. A comparison of the conversion of FFA obtained with both models (Cases I and II) was done. Figure 7 shows the experimental information and the result of both kinetic models. It can be seen that Case II presents a more accurate fitting. Downloaded from at Pennsylvania State University on March 6, of 6 International Journal of Low-Carbon Technologies 2010, 0, 1 6

5 Production of biodiesel from acid Figure 6. Experimental and calculated conversions of FFA. Dashed line area of the curve denotes the simulated values and filled symbols denote experimental data. Case II: initial FFA ¼ 10.68%; n ¼ 6.1:1; rpm ¼ 200; T ¼ 458C. Table 2. Kinetic parameters. Parameter Value Units k l/(mol s) E kj/mol k l/(mol s) E kj/mol k l/(mol s) E kj/mol k l/(mol s) E kj/mol k l/(mol s) E kj/mol k l/(mol s) E kj/mol k l/(kg s) E kj/mol k l/(kg s) E kj/mol K l 2 /mol 2 K l 2 /mol 2 K l/mol Figure 7. Comparison of the two models: continuous line, Case I; dashed line, Case II; bullets, experimental data. 4 CONCLUSIONS The kinetic reaction of the direct esterification was studied in the presence of triglycerides with a homogenous acid catalyst (sulfuric acid), obtaining good results. As a first step, the esterification reaction was studied, considering the triglycerides as inert because the transesterification reaction did not take place. A mechanism was established and the experimental data were fitted using a multiparametric non-linear routine. Satisfactory results were obtained for different reaction temperatures and amounts of catalyst (Figures 1 and 2). The laboratory experience showed that the transesterification reaction took place at the same time the esterification was studied. For this reason, a more complete mechanism was proposed involving both reactions. This model fitted the experimental data more accurately. This kinetic study is an important tool to study the reaction rate of the esterification and the transesterification reactions of an oil with high amount of FFAs. This kinetics is suitable for reactor design. REFERENCES [1] Schuchar U, Serchelim R, Vargas RM. Transesterification of vegetable oils: a review. J Braz Chem Soc 1998;9: [2] Fukuda H, Kondo A, Noda H. Biodiesel fuel production by transesterification of oils. J Biosci Bioeng 2001;92: [3] Srivastava A, Prasad R. Triglycerides-based diesel fuels. Renew Sust Energy Rev 2000;4: [4] Marchetti JM, Miguel VU, Errazu AF. Possible methods for biodiesel production. Renew Sust Energy Rev 2007;11: [5] Vicente G, Martinez M, Aracil J. Integrated biodiesel production: a comparison of different homogenous catalysts systems. Bioresour Technol 2004;92: [6] Di Serio M, Cozzolino M, Giordao M, et al. From homogeneous to heterogeneous catalysts in biodiesel production. Ind Eng Chem Res 2007;46: [7] Lopez DE, Goodwing JG, Jr., Bruce DA, et al. Transesterification of triacetin with methanol on solid acid and base catalysts. Appl Catal A Gen 2005;295: [8] Perira W, Close V, Patton W, et al. Transesterification with an anion-exchange resin. J Org Chem 1969;34: [9] dos Reis SCM, Lachter ER, Nascimento RSV, et al. Transesterification of Brazilian vegetable oils with methanol over ion-exchange resins. JAmOil Chem Soc 2005;82: [10] Lu PM, Li LH, Liu WW, et al. Biodiesel production from highly acidified oil through solid acid catalyst and plug flow reactor. Mod Chem Indus 2007;27:52 5. [11] Shibasaki-Kitakawa N, Honda H, Kuribayashi H, et al. Biodiesel production using anionic ion-exchange resin as heterogeneous catalyst. Bioresour Technol 2007;98: [12] Freedman B, Butterfield R, Pryde E. Transesterification kinetics of soybean oil. J Am Oil Chem Soc 1986;63: [13] Mittelbach M, Trathnigg B. Kinetics of alkaline catalyzed methanolisis of sunflower oil. Fat Sci Technol 1990;92: Downloaded from at Pennsylvania State University on March 6, 2016 International Journal of Low-Carbon Technologies 2010, 0, of 6

6 J. M. Marchetti et al. [14] Darnoko D, Cheryan M. Kinetics of palm oil transesterification in a batch reactor. J Am Oil Chem Soc 2000;77: [15] Knothe G, Krahl J, Van Gerpen J. The Biodiesel Handbook. AOCS Press, [16] Canakci M, Van Gerpen J. Biodiesel production via acid catalysis. Trans ASAE 1999;42: [17] Canakci M, Van Gerpen J. Biodiesel production from oils and fats with high free fatty acids. Trans ASAE 2001;44: [18] Zhang Y, Dubé MA, McLean DD, et al. Biodiesel production from waste cooking oil: 1. Process design and technological assessment. Bioresour Technol 2003;89:1 16. [19] Kusdiana D, Saka S. Methyl esterification of free fatty acids of rapeseed oil as treated in supercritical methanol. J Chem Eng Jpn 2001;34: [20] Saka S, Kusdiana D. Biodiesel fuel from rapeseed oil as prepared in supercritical methanol. Fuel 2001;80: [21] Sanchez F, Vasudevan PT. Enzyme catalyzed production of biodiesel from olive oil. Appl Biochem Biotechnol 2006;135:1 14. [22] Zhou M, Gilot B, Domenech S. Modelisation d un réacteur d estérification. Partie 1: Estérification de l acide oléique par le méthanol. Détermination des données thermodynamiques et cinétiques. Entropie 1984;120:3 10. [23] Goto S, Tagawa T, Yusoff A. Kinetics of the esterification of palmitic acid with isobutyl alcohol. Int J Chem Kinet 1991;23: [24] Bart HJ, Reidetschläger J, Schatka K, et al. Kinetics of esterification of levulinic acid with n-butanol by homogeneous catalysis. Ind Eng Chem Res 1994;33:21 5. [25] Rönnback R, Salmi T, Vuori A, et al. Development of a kinetic model for the esterification of acetic acid with methanol in the presence of a homogeneous acid catalyst. Chem Eng Sci 1997;52: [26] Lee MJ, Wu HT, Kang CH, et al. Kinetics of catalytic esterification of acetic acid with amyl alcohol over amberlyst 15. J Chem Eng Jpn 2001;34: [27] Gangadwala J, Mankar S, Mahaani S, et al. Esterification of acetic acid with butanol in the presence of ion-exchange resins as catalyst. Ind Eng Chem Res 2006;42: [28] Teo HTR, Saha B. Heterogeneous catalyzed esterification of acetic acid with isoamyl alcohol: kinetic studies. J Catal 2004;228: [29] Zhang Y, Ma L, Yang J. Kinetics of esterification of lactic acid with ethanol catalyzed by cation exchange resins. React Funct Polym 2004;61: [30] Sendzikiene E, Makareviciene V, Janulis P, et al. Kinetics of free fatty acids esterification with methanol in the production of biodiesel fuel. Eur J Lipid Sci Technol 2004;106: [31] Berrios M, Siles J, Martín MA, et al. A kinetic study of the esterification of free fatty acids FFA in sunflower oil. Fuel 2007;86: [32] Tesser R, Di Serio M, Guida M, et al. Kinetic of oleic acid esterification with methanol in the presence of triglycerides. Ing Eng Chem Res 2005;44: [33] AOCS method Ca 5a-40,1. [34] Marchetti JM. Technological Alternatives for Biodiesel production. Ph.D. Thesis. Universidad Nacional del Sur [35] Marchetti JM, Errazu AF. Esterification of free fatty acids using sulfuric acid as catalyst in the presence of triglycerides. Biomass & Bioenergy 2008;32: [36] Marquar DW. An algorithm for least-squares estimation of nonlinear parameters. Indust Appl Math 1963;11: [37] Streitwieser A, Heathcock CH, Kosower EM. Introduction to Organic Chemistry, 4th edn, Prentice Hall. ISBN: , Downloaded from at Pennsylvania State University on March 6, of 6 International Journal of Low-Carbon Technologies 2010, 0, 1 6